Convalescent plasma transfusion for immunocompromised viremic patients with COVID‐19: A retrospective multicenter study

This study aims to assess the safety, virological, and clinical outcomes of convalescent plasma transfusion (CPT) in immunocompromised patients hospitalized for coronavirus disease 2019 (COVID‐19). We conducted a retrospective multicenter cohort study that included all immunosuppressed patients with COVID‐19 and RNAemia from May 2020 to March 2023 treated with CPT. We included 81 patients with hematological malignancies (HM), transplants, or autoimmune diseases (69% treated with anti‐CD20). Sixty patients (74%) were vaccinated, and 14 had pre‐CPT serology >264 BAU/mL. The median delay between symptom onset and CPT was 23 days [13−31]. At D7 post‐CPT, plasma PCR was negative in 43/64 patients (67.2%), and serology became positive in 25/30 patients (82%). Post‐CPT positive serology was associated with RNAemia negativity (p < 0.001). The overall mortality rate at D28 was 26%, being higher in patients with non‐B‐cell HM (62%) than with B‐cell HM (25%) or with no HM (11%) (p = 0.02). Patients receiving anti‐CD20 without chemotherapy had the lowest mortality rate (8%). Positive RNAemia at D7 was associated with mortality at D28 in univariate analysis (HR: 3.05 [1.14−8.19]). Eight patients had adverse events, two of which were severe but transient. Our findings suggest that CPT can abolish RNAemia and ameliorate the clinical course in immunocompromised patients with COVID‐19.


| INTRODUCTION
In December 2019, a new coronavirus was identified in Wuhan, China.Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), the cause of coronavirus disease 2019 (COVID- 19), spread rapidly worldwide and led to an unprecedented health crisis.
Immunocompromised patients, who are unable to benefit from the protection provided by vaccination, have a higher risk of severe or protracted COVID-19. 1 Such is the case for patients with hematological malignancies (HM), especially those treated with anti-CD20; indeed, a meta-analysis indicated that such patients have a 32% fatality rate. 2 However, to date, there is no recommended treatment strategy for this population. 35][6][7] Direct antivirals are used for COVID-19 treatment but have low clinical efficacy 8 and carry the risk of drug-drug interaction (nirmatrelvir−ritonavir) in this heavily treated population.6][7] Monoclonal antibodies (mAbs) were effective in seronegative patients, but the loss of efficacy against SARS-CoV-2 subvariants is a concern. 9nvalescent plasma therapy (CPT) relies on passive immunotherapy as a curative treatment for COVID-19 and involves transfusion of plasma from immunized donors.The mechanism of action is dependent on polyclonal anti-SARS-CoV-2 antibodies targeting the spike protein, which exert a direct antiviral effect (including interference with the receptorbinding domain), and promote viral clearance via immunoglobulinmediated neutralization and other mechanisms (i.e., phagocytosis, cytotoxicity, and complement activation). 10Moreover, anti-inflammatory cytokines, clotting factors, and natural antibodies are obtained via transfusion, potentially improving inflammation. 11though several trials showed no benefit in the general population, 12,13 a recent randomized control trial showed better survival among intubated unvaccinated immunocompetent patients in the ICU when administered early after intubation. 14The trial was conducted early in the pandemic, among patients who did not benefit from antivirals or immunomodulators.6][17][18][19] However, these studies did not evaluate a virological endpoint, including SARS-CoV-2 in the blood (RNAemia), to assess the ability of CPT to enhance viral clearance.
We report the results of a retrospective multicenter observational cohort study of the clinical, virological, and safety outcomes of CPT in immunocompromised patients hospitalized for moderate to critical COVID-19 with RNAemia.Plasma was collected from convalescent donors (moderate COVID-19 1−6 months before donation) who were treated according to the Intercept ® Amotosalen technique.Patients provided written informed consent and received an average of 4 units of CPT (i.e., around 800 mL), in two or four transfusions at 24 h intervals.

| Study design
Convalescent plasma titers were 80 to >7000 BAU/mL (high-titer CP) during the study period.

| Data collection
The following clinical parameters were collected at baseline, during hospitalization, and follow-up: age, sex, body mass index, comorbidities, 2011 revised Charlson score, 20  binding antibody units/mL, as recommended by the WHO.A titer of >264 BAU/mL was considered positive. 22

| Outcomes and statistical analysis
The primary outcome was survival at D28, defined as the time between CPT administration and all-cause death.The secondary outcomes were clinical and biological evolution between CPT (D1) and D28, and the adverse effects of CPT.
Survival curves were plotted by the Kaplan−Meier method against predefined subgroups and compared by log-rank tests.To identify factors associated with all-cause death, Cox models were used to estimate hazard ratios (HR) and 95% confidence intervals (CIs).HRs for each factor of interest were estimated crude (univariate) and adjusted for the following variables (multivariate): age, sex, WHO score at inclusion (≥ or <6), and type of immunosuppression (B-cell HM, other HM, or no HM).
To identify factors associated with the persistence of viremia at D7 of CPT administration, only patients with an available blood PCR result at D7 were analyzed.Logistic regression models were used to estimate odds ratios (ORs) and 95% CIs for each factor of interest, crude (univariate), and adjusted for predefined variables (multivariate).Other exploratory analyses were conducted using the χ 2 or Fisher test (qualitative variables) and the Wilcoxon test (quantitative variables).
A two-tailed p-value < 0.05 was considered indicative of statistical significance.The proportionality of the instantaneous risk of death in the Cox models was verified by the Schoenfeld residual method.The linearity of quantitative variables was verified using cubic splines.No imputation of missing data was performed, given the small number of missing data.
Analyses were carried out using R studio (version 1.1.463;R Development Core Team) and the survival, survey, and cobalt packages.

| Factors associated with all-cause mortality
The overall mortality rate at D28 was higher in patients with non-Bcell HM (62%) than with B-cell HM (25%) or no hemopathy (11%) (p = 0.02).Patients on anti-CD20 without other chemotherapy had the lowest mortality rate (8%) (Table 3).Survival curves according to WHO score and type of immunosuppression are shown in Figure 2.
In univariate and multivariate analyses, severe disease (WHO score ≥6) at D0 (before CPT) was associated with a higher risk of   c Charlson score revised version (Quan et al. 20 ).
d Among those alive and with available PCR at D7 (n = 64).
e Among those living and with available serology at D6−D8 who were previously negative for mAbs (n = 34).

| RNAemia evolution after CPT
At D7, among 65 patients with available plasma PCR, 21 had persistent RNAemia.Clinical severity (WHO ≥ 6) at D0 was significantly associated with persistent RNAemia at D7 in a multivariate analysis (HR: 6.6 [CI: 1.84−23.7]).After April 12, 2022, RNAemia was monitored using the Cobas ® SARS-CoV-2 Kit.For 24 patients with at least one available PCR between D3 and D28 using the Cobas kit, the evolution of viremia using E and ORF-1 as targets is shown in Figure 3. Serology monitoring among patients without previous mAbs showed initial seroconversion followed by a rapid decrease during follow-up (Figure 4).

| DISCUSSION
We report an overall mortality rate of 25.9% at D28 after CPT in immunocompromised patients with COVID-19.The overall mortality rate was significantly higher in patients with non-B-cell HM (62%) and B-cell HM (25%) than with no HM (11%).Patients on anti-CD20 without chemotherapy had the lowest mortality rate (8%).After CPT, 67% of patients lost RNAemia, and 82% seroconverted at D7.
Persistent RNAemia at D7 was strongly associated with baseline clinical severity (WHO ≥ 6) and mortality at D28.
This study is the first of RNAemia and serology in immunocompromised patients receiving CPT.Our results are consistent with prior reports on the association of SARS-CoV-2 viremia and its persistence with severity and mortality, 24 and of that between seroconversion and viral clearance. 25This strong association between RNAemia and clinical severity emphasizes the importance CPT reportedly benefits immunocompromised patients with acute or protracted COVID-19, particularly those with humoral immunodepression. 15,16,18A meta-analysis of immunocompromised patients receiving CPT or standard-of-care showed a lower mortality in the CPT group. 17 therefore, analysis on serology is at risk of selection bias.Convalescent plasma titers were 80 to >7000 BAU/mL during the study period, but transfusion of 4 CPT units potentially reduces the risk of an insufficient antibody titer. 27The seroconversion threshold (>264 BAU/mL), which is commonly used, may not be relevant for Omicron variants of SARS-CoV-2. 28ven the high mutation rate and the ineffectiveness of mAbs, CPT (if collected recently) reflects the current circulating variants and the polyclonal immune response of several donors.Our data suggest its safety, as did prior work. 29r findings suggest viremic seronegative immunocompromised patients, hospitalized with noncritical and/or persistent COVID-19, as the target population.For ICU patients, CPT should be administered as early as possible to abolish RNAemia before clinical worsening.RNAemia and serology monitoring enable assessment of efficacy.Our findings support a link between seroconversion and RNA undetectability, but serologic positivity was not prolonged, possibly due to antibody consumption.This suggests the need for higher titer CPT, which has increased efficacy against the most recent Omicron variants. 30Repeated CPT administration could be effective in patients with persistent RNAemia.Comparative studies are necessary to confirm this hypothesis and formulate a treatment strategy.
This observational retrospective multicenter cohort study was conducted in eight hospitals in six cities in the Nouvelle-Aquitaine region of France.On April 29, 2020, the French national authorities authorized the collection and use of convalescent plasma as part of a therapeutic-use protocol.Following this decision, a multidisciplinary team meeting (CPT team) was organized in St-André University Hospital, Bordeaux.The meeting involved intensivists, infectious disease specialists, and French regional blood establishment referral physicians, with a view to allocating plasma in the Nouvelle-Aquitaine region.Between May 2020 and March 2023, the French Blood Establishment carried out 125 CPT deliveries, which were validated by the CPT team.This study included only immunosuppressed patients (patients with HM, autoimmune disease, solid organ transplantation, or active solid cancer; or receiving chemotherapy or other immunosuppressive treatment), hospitalized with COVID-19 (WHO score ≥4) with positive RNAemia before CPT.
immunosuppressive disease and treatment, COVID-19 vaccination, and pre-CPT treatment.Clinical status was evaluated by calculating the 10-point World Health Organization (WHO) clinical score. 21The WHO score, biological parameters, and virological parameters (SARS-CoV-2 RNAemia and serology) were collected at baseline (D0, before CPT), D3 (D2−D4), D7 (D5−D9), D14 (D12−D16), and D28 (D26−D30).Chest computed tomography was conducted to quantify the severity of lung damage at baseline.Among patients in intensive care units (ICU), the IGS2 and SOFA scores were determined at baseline.The causes of clinical worsening and deaths and the need for additional treatments were recorded.RNAemia was evaluated using the COVID-19 R-GENE (BioMérieux) until April 2022, and the Cobas ® SARS-CoV-2 Kit (Roche Diagnostics) thereafter.Thus, the viral targets differed over time (nucleocapsid [N], RNA-dependent RNA polymerase [RdRp], and Envelope [E] for R-GENE and E and ORF-1 for Cobas).Cycle threshold (CT) values ≤ 40 were considered positive results.SARS-CoV-2 serology was reported as
Between D1 and D28, 83.5% (66/79) of patients with available PCR did not have RNAemia.Nine of the 10 patients with persistent RNAemia after D7 (D14−D28) died.Six patients (all with WHO scores ≥6 at D0) had a second CPT (dose 2−8 units) with a median delay of 7 days (4−10 days) because of RNAemia persistence.RNAemia became undetectable in five of the six patients, one of whom died.The patient with persistent RNAemia died.Eight patients experienced a side-effect of CPT according to the definitions of the French National Authority for Health 23 : seven cardiovascular events (hypertension, dyspnea), and one cutaneous eruption.Two were severe and required increased oxygen and intubation; these occurred in patients in the ICU with multiple comorbidities (one cardiac transplant with heart failure and severe coinfections, one with bi-pulmonary transplant and severe chronic renal failure) and were rapidly resolved.

F I G U R E 2
Survival curves according to WHO score (A), type of immunosuppression (B), and RNAemia (C). of rapid viral neutralization.The long time from symptoms onset to CPT (median 23 days) confirms the longer time of viral replication in immunocompromised patients with often prolonged phase of moderate symptoms before worsening.In this study, RNAemia C tvalues increased until lack of detection after CPT.Nasopharyngeal PCR swabs monitoring, with sometimes long-lasting positivity despite clinical resolution, was less effective and related to clinical evolution than RNAemia.26 In association with CPT, antiviral treatment can optimize viral shedding.In this study, patients receiving nirmatrelvir/ritonavir showed less clinical worsening and nonsignificantly lower mortality.This could be because of a lack of power or the fact that patients on medical wards more frequently received nirmatrelvir/ritonavir than those in the ICU (37% vs. 14%).This treatment was authorized in France in January 2022; hence it was evaluated only for patients in the last year of the study and.Before CPT, 92% of patients received anti-inflammatory drugs such as corticosteroids and tocilizumab.These therapies could suppress the cytokine storm induced by COVID-19 but not RNAemia.Clearance is fundamental for immunocompromised patients, given that low viral loads can lead to symptomatic relapse.This study has several strengths.It was a multicenter study and evaluated clinical outcomes in hospitalized severely immunocompromised patients, who do not fully benefit from vaccination.We developed a standardized procedure for collecting clinical and biological data in multidisciplinary team meetings.This is the first study to monitor RNAemia in patients receiving CPT.F I G U R E 3 Evolution of RNAemia after CPT.*Among patients with at least two repeated C t values using the same PCR technique (E and ORF1 as targets) from April 2022 to March 2023.CPT, convalescent plasma transfusion.This study also has limitations.It was a retrospective, noncomparative study, and therefore it is possible that clinical improvement and viral clearance were a result of other treatments or the natural evolution of the infection.The number of patients was insufficient to envisage a precise target population.Missing data for clinical characteristics were scarce, but numerous for serology;

5 |
CONCLUSIONOur findings suggest that CPT reduces RNAemia and ameliorates the clinical course in viremic seronegative immunocompromised patients with COVID-19, particularly those with humoral immunosuppression.